[arin-ppml] Fairness of banning IPv4 allocations to somecategoryof organization

[Geoff Huston]

On 14/10/2009, at 2:35 AM, Ted Mittelstaedt wrote:

> Milton L Mueller wrote:
>>> [1] Section 3.3., "The Transferable Address Block Lease (TABL)," in
>>> Mueller, "Economic Factors in the Allocation of IP Addresses." http://www.itu.int/net/ITU-T/ipv6/ 
>>>  http://www.itu.int/dms_pub/itu-t/.../T3B020000020003PDFE.pdf
>> Tom Thanks for calling attention to the ITU report, I was actually
>> unaware that it had finally been released publicly.
>> Understand what this report is. ITU was authorized by its member
>> states to investigate its possible role in IPv6 address allocation.
>> In connection with that mandate, it commissioned two reports. One,
>> performed by the NAV6 center in Malaysia, investigated the
>> possibility of doing IPv6 allocation by having the ITU delegate
>> address blocks to Country Internet Registries (CIRs). The other,
>> performed by me, investigated the use of “market forces” as an
>> alternative to both ITU delegation and RIR needs assessment in v6
>> allocation. As was written in the report, “The purpose of [the
>> report] is not necessarily to advocate market-oriented policies, but
>> simply to gain a better understanding of what options exist and what
>> implications they might have for [IPv6] address management.”
>> As an organization whose members are nation-states, ITU obviously
>> favors the CIR approach. ITU should be credited, however, with some
>> degree of impartiality and integrity for its willingness to support
>> an exploration of other alternatives. (It is worth noting that its
>> staff's reaction to my critique of needs-based allocation was similar
>> to yours.)
>
> There are no other realistic alternatives.  ITU supported an
> exploration because it's obvious that the opponents of a CIR  
> approach -
> conservatives like you - are all favoring a Wild West, sale to the
> highest bidder market for IP addresses, and they might as well
> document how terrible that would be for all to see.
>
> You did your job admirably - but you still fail to grasp or respond
> to the underlying critique of a market-oriented policy when it comes
> to IP assignments, namely that by turning the Internet into a
> rich man's playground, it will destroy the fundamental usefulness of
> it.



I happen to agree entirely with Ted in this comment that this latest  
nonsense from Milton and his ITU cronies represents a fundamental  
threat to the viability of the network by promoting an address  
distribution architecture that is unroutable at a global scale. This  
approach that Milton has been pedalling to the ITU-T for more than  
five years now is simply poorly conceived and poorly thought through,  
and it fosters outcomes that will destroy the utility of the Internet  
as Ted comments above. Its dismaying to see the ITU picking up on this  
- I truly thought that there were at least some folk in the ITU-T who  
were able to exercise some adult supervision in this area of activity.  
Evidently not.

When I first heard of this work from Milton back in 2005 I was  
prompted to collaborate with Paul Wilson and write this item in  
response: http://www.potaroo.net/ispcol/2005-04/compete.html

Nothing has changed over the past 5 years. When I read the revised ITU- 
published documents last week I could see anything new in it - it  
contains the same faulty thinking, the same poor research and the same  
warped attempts at logic based on a pretty comprehensive  
misunderstanding of the way in which a self-regulatory framework  
operates in this environment of address distribution and routing for  
the Internet.

>
>> Contrary to your representation, the report takes full account of the
>> relationship between address allocation and routing, and indeed
>> constitutes one of the first attempts to understand that relationship
>> from the standpoint of institutional economics theory.

That claim is of course arrogant nonsense!


>> Obviously, a
>> lot more sophisticated analysis could be and needs to be done (and
>> not just apologetics for the status quo such as yours).

or ill-concieved and poorly informed rhetoric cloaked in the format of  
a study as in the case of these ITU-published documents?


>> The report
>> refers repeatedly to the “interdependence of routing and addressing,”
>> and says in Section 1.1, “The policy problem we are faced with is
>> not, in fact, one of efficiently allocating IP addresses per se. The
>> overarching problem is actually the efficiency and scalability of
>> Internet routing. The possession of IP addresses is significant and
>> valuable only insofar as they can be used to route packets on the
>> public Internet.” Please read section 1.3 for further analysis.
>
> You cannot expect the mindless market to distribute IP addressing to
> enhance Internet access for everyone.  Introducing the market merely
> changes the IP address distribution to favor the wealthy.

I suspect that there are a range of possible outcomes  and the  
scenario you paint here Ted is one possible such outcome. Other  
outcomes tend towards a "race to the bottom" in promoting address  
policies that emphasise "address portability" for consumers and  
unleash onto an unsuspecting and ill-equipped routing system a few  
hundred million independently routed /48s. None of these outcomes fill  
me with any confidence at all that the authors of these ITU papers  
have the slightest clue of what they are talking about in terms of  
their impact on network operators, on consumers and on the network's  
viability as a whole. If we really want to submarine any hope  
whatsoever of a transition to IPv6 this form of meddling in the  
address distribution framework is about as good a negative  
intervention as one could possibly imagine! It seems ironic to me that  
this proposal has been funded by the ITU, who has always had a deep  
and abiding interest in the welfare of the developing economies in our  
world. If there was just one aspect of Ipv6 that truly makes changes  
the environment, its a change in the availability of addresses in  
IPv6. At last we have the ability to not just ensure that there is one  
address for every person in this planet but one address for every  
device on this planet now and for a century or more to come. When ARIN  
discussed this at an IPv6 roundtable back in 2005 David Conrad  
presented the wonderful analogy that if an address was the same as a  
single grain of sand then IPv6 does not just represent the 90 mile  
beach, or the entire sahara or even the entire plant, but 300 million  
planets each the size of the entire earth. But what has NOT changed in  
the slightest  is routing technology, We still have the same ceiling  
in routing that has managed to climb from a few tens of thousands of  
routing slots twenty years ago to a few millions of slots today. So  
the challenge here is one of self-imposed constraint in address  
distribution policies that strike rational balances between utility  
and versatility, between consumer and service provider, between global  
and local considerations. None of the ITU published material appears  
to address this underlying constraint, yet this precise constraint and  
this same effort of  attempting to strike an appropriate balance has  
been at the heart of the address policy work I've seen in the RIR  
forums over the past decade and longer.

I fail to see why a great leap backward into the morass of squabbling  
nation states, national monopolies, and the bizarre form of attempting  
to make rational technology decisions via voting across national  
delegations will work any better than the efforts of the ITU to push  
OSI two decades ago. If the Internet truly is a product of the  
progressive wave of deregulation in our industry and if it relied on  
the dismembering of such insular and inefficient national fiefdoms  
then I cannot see why we should give credence to such ill founded  
arguments to rebuild a now irrelevant and unneeded past.



>
>> Your attempt to criticize the report’s approach to the impact of
>> TABLs on routing misses the mark. You commit a simple logical error,
>> or perhaps didn’t actually read the proposal details, or maybe are
>> simply misrepresenting what was said, it is hard to tell. You imply
>> that address blocks would be auctioned when in fact, the report
>> considers and rejects auctions as a method for initial allocation of
>> ipv6 address blocks. See section 3.2. It proposes, instead, that RIRs
>> use fees geared to size of address blocks to allocate resources. The
>> TABL proposal specifies that blocks so allocated cannot be
>> disaggregated and parcelled out to other ASs. By saying that the
>> proposal "would not make the current system any worse or any better
>> with respect to routing table growth" my point was that for routing
>> purposes it matters not at all whether the initial allocation of ipv6
>> address blocks is based on “need” as defined by RIRs or on the basis
>> of fees, as suggested in the report – what matters is the size of the
>> blocks and the degree to which they are disaggregated. Can you
>> dispute that?

The ITU report proposes a market rationalist approach to address  
distributing by advocating blocks to countries and allowing countries  
to develop their own national address allocation frameworks? You are  
really calling this a "market based approach" Milton? You've made some  
rather strange claims in the past mate, and this one would have to  
rank right up there with them.

   Geoff

   Who is writing as myself as an individual here.